最終更新日:2026-10-28 14:17:56
Reading time: About 7 minutes
Prerequisites: The Flood Shield 2.0 service has been activated, and at least one non-website (TCP/UDP) service has been onboarded.
Usage Statistics lets you view the bandwidth, traffic, and concurrent connections of your non-website services within a specified time range, helping you assess current traffic load and resource consumption.
This page is under Non-Web Analysis and only covers TCP/UDP services onboarded through port forwarding. For usage data of HTTP/HTTPS website services, go to Web Analysis.
Path: Non-Web Analysis > Usage Statistics
| Filter | Description |
|---|---|
| Time Range | The time period to query; a more precise range gives more targeted results |
| Control Group | Filter by control group, so you can view the usage of each control group separately when you run multiple services |
| Domain | Select multiple domains, or select all to view data across every domain |
The Bandwidth & Traffic tab shows the bandwidth and traffic generated at the edge nodes. By default, it plots three curves at once:
| Curve | Description |
|---|---|
| Edge Upstream | Bandwidth in the client → edge node direction |
| Edge Downstream | Bandwidth in the edge node → client direction |
| Total Bandwidth | The sum of edge upstream and edge downstream, used to assess overall bandwidth consumption |
If you need to view a single direction on its own, use the bandwidth type filter to keep only that curve.
If bandwidth spikes, first compare it with the Concurrent Connections curve for the same period: if bandwidth and concurrent connections rise together, this usually reflects organic business growth; if bandwidth rises while concurrent connections stay roughly flat, it is more likely that traffic on individual connections has been amplified — for example, by large file transfers or traffic abuse.
The Concurrent Connections tab shows the number of connections established simultaneously within the query period. You can filter by IPv4 or IPv6 to view concurrent connections for each network-layer protocol separately.
Concurrent connections and bandwidth are tracked independently. If bandwidth stays low but concurrent connections remain persistently high, this usually indicates a large number of short-lived or idle connections. Check Analysis & Logs > Attack Logs to further confirm the source.
The Region Distribution tab aggregates usage within the query period by visitor region. The table is sorted by peak bandwidth from highest to lowest, with each row showing:
| Field | Description |
|---|---|
| Peak Bandwidth | The highest bandwidth this region reached within the query period |
| Peak Time | The point in time when the peak bandwidth occurred |
| Total Traffic | The region’s cumulative traffic within the query period |
“Region” refers to where the end user is located, not the location of the origin server or edge node. Use the region filter to select the regions you want to view.
If one region’s share of traffic is markedly higher than others and does not match your expected business distribution, go to Non-Web Analysis > Top Rankings and use the Visitor tab to pinpoint the specific IP.